DC Field | Value | Language |
---|---|---|
dc.contributor.author | Park, Cheoneum | ko |
dc.contributor.author | Shin, Jamin | ko |
dc.contributor.author | Park, Sungjoon | ko |
dc.contributor.author | Lim, Joonho | ko |
dc.contributor.author | Lee, Changki | ko |
dc.date.accessioned | 2023-07-06T07:00:26Z | - |
dc.date.available | 2023-07-06T07:00:26Z | - |
dc.date.created | 2023-06-08 | - |
dc.date.issued | 2020-11 | - |
dc.identifier.citation | Findings of the Association for Computational Linguistics: EMNLP 2020, pp.2610 - 2624 | - |
dc.identifier.uri | http://hdl.handle.net/10203/310347 | - |
dc.description.abstract | Recently, end-to-end neural network-based approaches have shown significant improvements over traditional pipeline-based models in English coreference resolution. However, such advancements came at a cost of computational complexity and recent works have not focused on tackling this problem. Hence, in this paper, to cope with this issue, we propose BERT-SRU-based Pointer Networks that leverages the linguistic property of head-final languages. Applying this model to the Korean coreference resolution, we significantly reduce the coreference linking search space. Combining this with Ensemble Knowledge Distillation, we maintain state-of-the-art performance 66.9% of CoNLL F1 on ETRI test set while achieving 2x speedup (30 doc/sec) in document processing time. | - |
dc.language | English | - |
dc.publisher | Association for Computational Linguistics (ACL) | - |
dc.title | Fast end-to-end coreference resolution for Korean | - |
dc.type | Conference | - |
dc.identifier.scopusid | 2-s2.0-85118461741 | - |
dc.type.rims | CONF | - |
dc.citation.beginningpage | 2610 | - |
dc.citation.endingpage | 2624 | - |
dc.citation.publicationname | Findings of the Association for Computational Linguistics: EMNLP 2020 | - |
dc.identifier.conferencecountry | US | - |
dc.identifier.conferencelocation | Virtual | - |
dc.contributor.localauthor | Park, Sungjoon | - |
dc.contributor.nonIdAuthor | Park, Cheoneum | - |
dc.contributor.nonIdAuthor | Shin, Jamin | - |
dc.contributor.nonIdAuthor | Lim, Joonho | - |
dc.contributor.nonIdAuthor | Lee, Changki | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.